Boucher2010
CREATED: 201001261251 LINK: url:~/Modules/Literature/Boucher2010.pdf Title: Fast motif recognition via application of statistical thresholds
Approach
construct a graph (V = seq, W = hamming distance)
run MCL to find dense regions
previous method: use dynamic programming to check whether consensus exist
use weights of a set (sum of pairwise hamming distance) as an indicator whether a set of sequences has a consensus, seems to work well in practice
effective rejection sampling of pairwise bounded sets